Salmonella in low-moisture foods: chocolate, nuts, spices and milk powder

What is a low-moisture food?

The Codex Alimentarius defines low-moisture foods as those with a water activity (a_w) of 0.85 or less. In practice, many products are well below that value: chocolate, nuts and nut pastes, spices and dried herbs, cereals, cookies, milk powder and other dried ingredients. These products stay stable against spoilage for months or years, and precisely for that reason they were long considered low microbiological risk.

The experience of recent decades has disproved that idea. A systematic review commissioned by FAO and WHO identified 214 outbreaks and case reports associated with these foods, and Salmonella was implicated in 45% of them, accounting for most of the hospitalizations and deaths recorded. The general biology of the pathogen is covered in our guide on Salmonella in the food industry; here we focus on what changes when the matrix is dry.

It does not grow, but it survives: the key role of water activity

Gram-negative bacteria, including Salmonella, need an a_w above 0.93 to multiply. In chocolate or milk powder there is not enough available water for the population to increase, so the count does not grow over time. The problem is that it does not disappear quickly either: the bacterium slows its metabolism, accumulates protective solutes such as trehalose and enters a resistant state that keeps it alive.

The laboratory data are compelling. In storage studies, a Salmonella Eastbourne strain was still recoverable from inoculated chocolate after 19 months, and survival of more than 100 weeks has been documented for dried cells on plastic surfaces under refrigeration. For the quality manager, this has a direct consequence: contamination that happens today can reach the consumer at the end of a very long shelf life, with no change in the product's appearance, odor or taste.

More heat-resistant, too

Dehydration does not only prolong survival; it also increases heat tolerance. The Codex warns that microorganisms are more heat-resistant in matrices with reduced water activity, and that the degree of resistance varies with the specific ingredients. A treatment that comfortably inactivates Salmonella in liquid milk can fall short in a powder, a nut paste or a high-fat cocoa mass.

That is why roasting, steam treatments and other microbial reduction steps must be validated in the real matrix and under the product's real moisture conditions, not extrapolated from data obtained in high-moisture foods. The scientific literature agrees that the increase in resistance is general, but its magnitude varies from one product to another and must be assessed case by case.

Low infective dose in fatty matrices

In many foods it is estimated that more than 10⁵ Salmonella cells are needed to cause illness. In outbreaks linked to dry products, however, the amounts recovered from the implicated food were very small: in the order of 10 to 100 CFU in chocolate outbreaks, and fewer than an estimated 45 cells in a 1993 outbreak in Germany caused by paprika-seasoned potato chips, which affected about 1,000 people.

Why fat protects the bacterium

The most widely accepted hypothesis, also cited by the Codex, is that a fat-rich matrix such as chocolate or peanut butter protects the cells against stomach acid and allows them to reach the intestine alive. In addition, contamination is often unevenly distributed, in small clusters of cells, which can lead to underestimating the actual dose ingested. In practice, the criterion for these products is absence: no level of Salmonella can be considered tolerable.

Lessons from historical outbreaks

The list of outbreaks linked to dry foods spans almost every category in the sector:

• Milk powder: an outbreak in Trinidad in 1973 affected about 3,000 people.
• Spices: besides the paprika case in Germany, in 2009-2010 in the US 272 people fell ill from sausages made with contaminated imported black and red pepper.
• Nuts and nut pastes: a 2008-2009 outbreak in the US linked to peanut butter and peanut paste exceeded 700 cases and may have contributed to nine deaths.
• Chocolate: in 2022, EFSA and ECDC investigated a multi-country outbreak of monophasic Salmonella Typhimurium linked to chocolate products, with 324 confirmed and probable cases in Europe, mostly children under 10, and 41% hospitalized. The outbreak strains were found on the processing line of a fatty dairy ingredient at the plant.

The pattern repeats itself: contaminated ingredients, an insufficient microbial reduction step or later recontamination from the plant environment. If you work in sweets, the segment analysis is in our article on confectionery microbiology.

Water, the real enemy in a dry plant

According to the Codex, outbreak investigations indicate that the safety of these products depends fundamentally on controlling Salmonella in the processing environment. And in that environment, the critical factor is water. A crack or a hard-to-clean spot that stays dry is a contained problem; when water gets in, through a leak, condensation or poorly controlled washing, the bacterium can multiply there and become a niche that contaminates the product intermittently.

Three levels of cleaning

• Dry cleaning: the routine practice in areas with the highest hygiene requirements, using vacuums, brushes and dedicated tools, with no water.
• Controlled wet cleaning: minimal use of water only when essential, for example after an environmental finding, followed by complete drying and verification before production resumes.
• Wet cleaning: reserved for non-critical areas away from processing.

The FDA incorporated the same approach into its 2025 draft guidance for low-moisture ready-to-eat foods, which covers products such as chocolate, processed nuts, milk powder and ground spices, and includes recommendations on water control, sanitation programs, routine environmental monitoring and root cause investigation after a contamination event.

Product and environmental sampling

The Codex recommends an environmental monitoring program for products of known risk, with Salmonella as the main target organism and samples that include swabs, dust and product residues. Enterobacteriaceae are useful as a process hygiene indicator, but the Codex itself warns that even low levels do not guarantee the absence of Salmonella, so they do not replace testing for the pathogen.

For finished product, uneven distribution calls for statistical sampling plans with several units per batch, because a single negative result says little about the whole batch. In addition, cells that survive in dry matrices are often stressed, so the enrichment step is decisive for them to recover before detection.

How it is detected with molecular methods

Dry and fatty matrices pose their own challenges for the laboratory: the bacterial load is usually very low and the product's composition complicates sample preparation. Molecular methods shorten turnaround time compared with culture, provided they are validated in the specific matrix you analyze. Amplifying RNA instead of DNA, as explained in our article on RNA vs. DNA, focuses detection on viable cells.

Along these lines, TAAG's Elevia 1.1 Salmonella spp. uses the RNA-based AiGOR™ RT-qPCR technology, is AOAC-certified by extension of license No. 072601, with an AOAC scope that matches this type of product: 62% cocoa chocolate, cocoa liquor, skimmed milk powder, peanuts and enriched stainless steel surface samples. The matrices validated by TAAG also include hazelnut-type and milk-type filled chocolates, chocolate bar with peanuts, hazelnut and cocoa cream and animal feed pellets. Other nuts and spices do not appear on either list: their scope must be confirmed with TAAG.